Investigating the Performance of Meta-Exploratory Algorithms in Two-Dimensional Truss Optimization (Genetic Algorithm)

The optimization of truss structures, including the optimization of topology, shape and cross-sectional area (size) of truss members, has been the focus of various researchers over the past years. The aim of this study is to optimize the discrete and continuous size of two-dimensional trusses with fixed topology and shape. For this purpose, the cross-sectional area of all the members is selected as the design variables and minimizing the weight of the structure as the objective function, and the constraints of the problem include the restrictions related to the changes in the location of the nodes and the tension in the members, the permissible values of which are determined using the conditions of the problem. In this study, genetic algorithm is used to optimize the truss. OpenSees structural analysis software is used to analyze the structure and obtain member forces and displacement of nodes. This software is properly connected with genetic algorithm code and particle swarm algorithm prepared in MATLAB software. In this research, the optimization of 2 two-dimensional trusses, including a six-node ten-member truss and an eight-node fifteen-member truss, is investigated, and the results of this research are compared with the results of previous articles. The results of the comparison show that in discrete sizes, the plans obtained from the genetic algorithm are generally more economical than other plans.